Functional Testing: Dashboard
Overview
Use this page to review functional test run results, testcase stability, and engine capacity for a selected period. The Functional Testing dashboard summarizes every headless LoadGen Studio functional test run on your enrolled Windows engines: how many passed or failed, which testcases are flaky, and how much engine capacity is in use.
The dashboard sits downstream of the rest of the module: you enroll Windows engines on the Engines page and configure and run functional test packages, manually or on a schedule, before any data appears here. Every completed run feeds this dashboard's KPIs and charts, and from here you drill into run-level detail on the Run History page.
This page itself lets you choose a reporting period, preset or custom, and read KPI cards for pass, fail, unfinished, and flaky counts plus run time, alongside the engine and run slot capacity bar. It also shows pass-rate and verdict charts, unstable testcases, recent runs, and run-time trends, and lets you export a PDF or ask LoadGen AI about the period.
Key concepts
- engine - A Windows host enrolled to execute LoadGen Studio functional test runs headlessly. Engines must be online to consume run slots and produce the runs this dashboard summarizes. (related: Functional Testing: Engines)
- run slot - A unit of concurrent run capacity, drawn from your Functional Testing licences at 10 slots per licence. The capacity bar shows how many of your available slots are currently occupied. (related: LoadGen Cloud: Licensing)
- testcase - An individual check inside a functional test package that runs headlessly on an engine and returns its own verdict each time the package runs. (related: Functional Testing: Packages)
- verdict - The pass, fail, or other outcome recorded for a completed functional test run. The KPI cards and charts group runs by verdict for the selected period. (related: Functional Testing: Run History)
- flaky testcase - A testcase whose last ten completed runs include at least three verdict changes, flagged so you can investigate it before relying on it as a release gate. (related: Functional Testing: Packages)
When to use it
- Before trusting a release gate, to confirm the recent period's functional runs passed and no testcase is flaky.
- When a scheduled run appears stuck, to check Engines online and Run slots in use for a capacity shortage.
- When deciding whether a testcase is reliable enough to keep, using the unstable testcases detail lower on the page.
- When you need a shareable summary of a period's results for a stakeholder, using Period report (PDF).
- When the capacity bar reports no engines online, to confirm enrollment before troubleshooting individual runs.
- When comparing short-term and long-term stability, by switching between the preset time ranges.
Before you begin
- At least one Windows engine enrolled using an enrollment token created on the Functional Testing: Engines page; otherwise the dashboard shows Engines 0 of 0 online and zero run slots in use.
- One or more functional test packages configured and run, manually or on a schedule, via Functional Testing: Packages and Functional Testing: Schedules, so the dashboard has runs to summarize.
- Enough Functional Testing licences to cover the run slots your concurrent runs need, since each licence provides 10 run slots.
- A user account with access to the Functional Testing module (the captured session uses a GlobalAdmin role).
Common workflows
Check whether the last 14 days of functional testing are healthy
Persona: Load-test operator
- Open the Functional Testing dashboard.
- Expect: The page loads with the 14d period already selected, and the KPI cards, capacity bar, and charts populate for the last 14 days.
- Read the Runs passed, Failed runs, Unfinished runs, and Flaky testcases KPI cards.
- Expect: Each card shows a count for the period, for example Runs passed as an N of M ratio.
- Check the capacity bar for Engines online and Run slots in use.
- Expect: The bar shows how many enrolled engines are online and how many of the licensed run slots are occupied.
- Select a shorter preset such as 24h if a KPI looks unexpected.
- Expect: The KPI cards, capacity bar text, and charts reload for the new 24-hour window.
Verify: Runs passed shows a total matching your expected test volume for the period, Unfinished runs and Flaky testcases are at or near zero, and the capacity bar reports your engines online.
Enroll an engine when the dashboard shows no capacity
Persona: Admin
- Read the capacity bar message stating no engine is enrolled yet.
- Expect: The message explains that you create an enrollment token on the Engines page and run the enroll command on a Windows host.
- Select Manage engines.
- Expect: The Engines page opens, where you can create an enrollment token.
- Return to the Functional Testing dashboard and select Refresh.
- Expect: The capacity bar updates to show the newly enrolled engine online and its run slots available.
Verify: Engines online rises above 0 of 0, and Run slots in use reflects the new engine's capacity instead of 0 of 120.
Steps
Step 1
Open the Functional Testing dashboard from the left navigation.
Expected result: The dashboard loads showing the Runs passed, Failed runs, Unfinished runs, Flaky testcases, and Median run time KPI cards, the engine and run slot capacity bar, and the start of the Validation pass rate per day and Runs per day by verdict charts for the default 14d period. Because no engines are enrolled and no runs exist yet, every KPI reads zero and the capacity bar reports 0 of 0 engines online.
This is the starting view you use to judge the health of your functional testing before drilling into individual runs.
- The 14d time range button is selected by default, covering runs created from 2026-10-26 11:29 to 2026-11-09 11:29.
- Period report (PDF) and Refresh sit in the header for exporting or reloading the current period.
- Manage engines links to the Engines page when the capacity bar shows no engines online.
Column reference
| Column | What it shows | How to read it |
|---|---|---|
| Runs passed | Shows how many of the period's completed runs ended with a pass verdict, out of the total completed runs; a ratio well below the total signals a period-wide problem rather than one testcase. | Reads 'no run passed or failed yet' until the first run in the period completes; afterwards compare the ratio against the total to judge overall pass rate. |
| Failed runs | Counts testcases the dashboard treats as not passing for the period, as a single figure separate from the raw engine verdict. | A rising count without a matching drop in Runs passed points to a small number of repeatedly failing testcases worth checking individually. |
| Unfinished runs | Counts runs that were cancelled, timed out, interrupted, or ended in an engine error instead of reaching a verdict. | A nonzero value means some scheduled runs never produced a result; check engine health before trusting the other KPIs for the same period. |
| Flaky testcases | Counts testcases with at least three result changes across their last ten completed runs, regardless of which period you select. | Any nonzero value means at least one testcase is unreliable enough to investigate before using it as a release gate. |
| Median run time | Shows the middle run duration for the period, with the 95th percentile (p95) duration shown underneath so you can distinguish a typical run from a worst-case one. | A median far below the p95 means a small number of runs take much longer than normal; check those outliers before assuming the whole period is slow. |
| Engines online | Shows how many of your enrolled engines are currently online out of the total enrolled, which is the capacity you can actually use right now. | 0 of 0 with no engines enrolled means no functional test can run until you enroll one from Manage engines. |
| Run slots in use | Shows how many of your total available run slots, licences multiplied by 10, are currently occupied by running tests, telling you whether you have headroom for more concurrent runs. | If the first number approaches the second, new runs may queue until a slot frees up or you add Functional Testing licences. |
| Selected period range | Shows the exact start and end timestamp the current preset or custom selection resolves to, so you can confirm precisely which runs are counted. | Use this line to line up the dashboard's figures with an external report or an incident window down to the minute. |
Controls reference
Time range filter (1h, 4h, 8h, 24h, 7d, 14d, 30d)
Switches the reporting period the dashboard summarizes, from the last hour up to the last 30 days.
When to use: Shorten the period to isolate a recent incident, or lengthen it to judge longer-term pass-rate and stability trends.
Side effects: Read-only: reloads the KPI cards, capacity bar text, and both charts for the new window; no data is changed.
Custom
Lets you set a specific date and time range instead of one of the presets.
When to use: When you need results for an exact maintenance window or incident period that does not align with a preset.
Side effects: Read-only: reloads the dashboard for the chosen range; no data is changed.
Refresh
Reloads the dashboard's KPIs, capacity bar, and charts from current data without changing the selected period.
When to use: After enrolling an engine or completing a run, to see the dashboard reflect the latest results.
Side effects: Read-only: reloads displayed data only.
Period report (PDF)
Generates and downloads a PDF summary of the same KPIs and charts for the selected period.
When to use: When you need to share or archive the dashboard's results outside the platform.
Side effects: Read-only: produces a downloaded file; does not change any data.
Manage engines
Opens the Engines page where you create enrollment tokens and view enrolled engines.
When to use: When the capacity bar reports no engines online or you need more capacity.
Side effects: Navigation only; no data changes on this page.
Ask AI about this period
Opens an AI query scoped to the results of the currently selected period.
When to use: When you want a natural-language explanation of why verdicts or stability changed over the period.
Side effects: Read-only: starts an AI conversation; does not change dashboard data.
Tag count chips (unstable testcases)
Filters the unstable testcases table to only the testcases carrying the selected tag.
When to use: To separate testcases that just started failing from ones that have never passed.
Side effects: Read-only: filters the rows shown in that table only.
Verify the result
A healthy functional testing dashboard shows Runs passed with a total close to the full count of the period's runs, Failed runs and Unfinished runs at or near zero, and Flaky testcases at zero. The capacity bar reads Engines N of N online with at least one engine online, and Run slots in use sits comfortably below the total available slots. The Validation pass rate per day and Runs per day by verdict charts show a bar or point for each day a run was created in the period. If instead every KPI reads zero and the capacity bar shows 0 of 0 online with a message that no engine is enrolled, no engine has been enrolled or no functional test has run yet in the selected period; check the Engines page first.
Troubleshooting
- Symptom: All KPI cards read 0 and the capacity bar shows Engines 0 of 0 online. Fix: Enroll a Windows engine from the Engines page using an enrollment token, then select Refresh.
- Symptom: Run slots in use is close to the total shown after the licence count. Fix: Reduce concurrent scheduled runs, or add Functional Testing licences on the LoadGen Cloud: Licensing page.
- Symptom: Flaky testcases stays above zero across every period you check. Fix: Open the unstable testcases table, use the tag count chips to isolate the affected testcase, and review its recent runs before using it as a release gate.
- Symptom: Median run time shows a dash even though you expect completed runs. Fix: Widen the time range, for example from 1h to 14d, since the figure only appears once enough completed runs fall inside the selected period.
- Symptom: Selecting Period report (PDF) produces no visible download. Fix: Check your browser's download or pop-up blocker; the report generates from the period currently selected on the dashboard.
Related articles
- Prerequisite: Functional Testing: Engines - Enroll the Windows engines that execute the runs this dashboard summarizes
- Prerequisite: Functional Testing: Packages - Define the functional test packages and testcases whose verdicts appear here
- Prerequisite: Functional Testing: Schedules - Schedule recurring runs so the dashboard has a steady stream of results
- Next step: Functional Testing: Run History - Drill into individual run details behind the dashboard's KPIs and recent runs
- See also: LoadGen Cloud: Licensing - Check or add the Functional Testing licences that determine total run slots
- See also: LoadGen AI: Ask LoadGen AI - Ask a natural-language question about the selected period's results